The Hydro-Estimator technique to estimate precipitation from satellite was originally developed at the National Oceanic and Atmospheric Administration / National Environmental Satellite, Data, and Information Service (NOAA / NESDIS). In Argentina, a local version of this technique became again operational at the National Meteorological Service from September 2013. It is based on thermal infrared channel images from GOES-13 satellite and forecasted meteorological variables from ETA model outputs. Rain and no rain pixels are separated according to a temperature value built with the mean and standard deviation of the brightness temperature at an area centered on the pixel being analyzed. The clouds in a given pixelproduce precipitation if they have colder tops than the average of the surrounding pixels. The precipitation rate is adjusted considering data from relative humidity and precipitable water of numerical forecast models. On the website of SMN, precipitation rain rate maps are produced and uploaded to internet every half hour, and accumulated precipitation maps every 6, 12 and 24 hours, with a spatial resolution of 4 km. Currently, monitoring and the application of different changes in the algorithm are being performed in order to improve the estimate. The changes included a version associated with filtering cirrus clouds, as well as a comparison with a version that does not discriminate clouds clusters. A validation is performed every 24 hours for this product and test versions, for a period of one complete year of data available. This integral evaluation is made with different statistical analyses locally and in the complete network of meteorological stations available in real time. In southern South America, where different precipitation regimes are present, it was observed in daily validation over the available months, that Hydro-Estimator tends to underestimate in weak precipitation events, as shown by Salio et al. (2015), but a significant overestimation appears in more intense precipitation events.